This study essentially seeks to assess the reconcilability of two themes of sustainable agriculture as a means to attenuating the sustainable agriculture impasse. It responds to the following question: relative to apparently pro-modernization international production studies, which element of pro-indigenous-knowledge case studies is atypical: the micro-environments studied or the methodologies deployed? In terms of methodology, the study applies the method of seemingly unrelated regression to a panel dataset of 124 countries mainly over the period 1981-2005 to estimate meta-production functions. The estimated coefficients are then used to evaluate series of values of the marginal physical/value productivity of agricultural land for each country. These values are used as indicators of agricultural resilience. The results of the analysis, which prove to be robust to the measure of agricultural output among other things, suggest that the resilience of national agricultural systems falls with increasing use of external inputs. It is demonstrated that this finding implies that agricultural resilience is eroded when Western science and technology is introduced from the angle of trying to replace rather than foster indigenous knowledge based systems. The contribution of this thesis has several implications for development and research policy. In general, economic analysis may need to (more fully) embrace the challenge of reworking its methods and approaches of research not least by opening up to alternative realities. In terms of development policy, the study unsettles the primacy of Western development, and reinforces evidence to the effect that sustainable agriculture might require enhancing, rather than nihilistically replacing, indigenous knowledges.